Risk factors for Loans in Chargedoff/Defaulted Status¶

by Christopher Hu¶

Investigation Overview¶

In this investigation, I want to study what are the main risk factors that will cause Prosper loans to be in Chargedoff/Default status

Dataset Overview¶

The dataset contains 84853 after-072009 Prosper Loans that include loan amount, borrower rate (or interest rate), current loan status, borrow income, income range, borrower state, Employment Status, Occupation, Prosper Rating, ... etc (total 81 variables) for each loan. Categorical factors such as EmploymentStatus, IncomeRange, ProsperRating_Alpha (CreditScoreRangeLower), ListingCategories, CreditScoreRangeLower, LoanOriginationYear, Occupation are used to study their relationship with LoanStatus and hence to predict the risk factors for the loan

NOTE: please click the legend of the plot to filter out Chargedoff/Defaulted Loans¶

Note that the above cells have been set as "Skip"-type slides. That means that when the notebook is rendered as http slides, those cells won't show up.

EmploymentStatus vs LoanStatus¶

People without any job (Not employed) has the highest percentage of the Chargedoff/Defaulted status (25%) while People with jobs (Employed) has the lowest (6.2%). People retired or having part-time job has also very high percentage of loans in Chargedoff/Default status

IncomeRange vs LoanStatus¶

Two groups of IncomeRange have the hightest Chargedoff and default loans: \$0 and Not employed (26.6% and 25.1% respectively). While IncomeRange > \\$100,000+ has the lowest chargedoff and defaulted loans (total 3.8%). This match the result from EmploymentStatus vs LoanStatus stacked barplots

# LoanOriginationYear vs LoanStatus¶

It is interesting to see that year 2013 and 2014, there is either low or no loan in chargedoff and defaulted status. This could be explained by the good economy in 2013 and 2014. Both 2013 and 2014 the chargedoff and defaulted loan rate are very low. Please read the following reference

  • https://obamawhitehouse.archives.gov/blog/2013/12/19/economy-2013#slides
  • https://obamawhitehouse.archives.gov/blog/2014/12/16/economy-2014

Occupation vs LoanStatus¶

It is very interesing to see that college sophomo, Freshman, and junior students have the top three highest proportion to be either get their loans into chargedoff and defaulted status.

LoanStatus vs EmploymentStatus and ProsperRating_Alpha¶

Chargedoff vs EmploymentStatus and ProsperRating_Alpha¶

Regardless of EmploymentStatus, the percentage(>80%) of the people who gets their loan into Chargedoff/Defaulted status are having a C or lower ProsperRating_Alpha letter (C, D, E, HR)

LoanStatus vs EmploymentStatus and ProsperRating_Alpha¶

Defaulted vs EmploymentStatus and ProsperRating_Alpha¶

Regardless of EmploymentStatus, the percentage(>80%) of the people who gets their loan into Chargedoff/Defaulted status are having a C or lower ProsperRating_Alpha letter (C, D, E, HR)

Conclusion:¶

Low Credit score, Low income, unemployment, and College Students are high risk factors for Prosper loans get into Chargedoff/Defaulted status. Good economic situation can minimize jobless people and hence help to reduce significant amount of chargedoff and defaulted loans.

Generate Slideshow: Once you're ready to generate your slideshow, use the jupyter nbconvert command to generate the HTML slide show. . From the terminal or command line, use the following expression.

This should open a tab in your web browser where you can scroll through your presentation. Sub-slides can be accessed by pressing 'down' when viewing its parent slide. Make sure you remove all of the quote-formatted guide notes like this one before you finish your presentation! At last, you can stop the Kernel.